Two placements, four rooms, four different months.

Why the pair

Equity says what was produced; drawdown says what it cost.

An equity placement is a statement about output with no price attached. The system could have reached P70 by operating exactly as modelled, or by consuming risk the model never authorised, and the equity band is identical in both cases. The drawdown placement supplies the missing half — the risk actually spent producing that output, located against what governed futures spend. Only the pair describes a transaction. This is the same principle as the conversion rows in the comparison stack, applied at the level of the two headline distributions, and it is why the standard states flatly that a strong equity result with abnormal drawdown is not clean alpha.

Room one

Strong equity, normal drawdown — clean, and the rarest of the four.

Above the centerline with risk behaviour inside its bands is the only room where the word outperformance survives contact with the cross-checks. The playbook's response is deliberately unexciting: continue the normal workflow, and document which branches, regimes, or throttle states contributed. Documentation is the whole action, and it is the action most often skipped because a good month feels self-explanatory. It is not. A clean month is the one period where the system's favourable behaviour can be attributed while the evidence is fresh, and the attribution is what makes the next clean month interpretable rather than merely pleasant.

Room two

Strong equity, adverse drawdown — fast but risky, and audited.

This is the room the pairing exists to expose. Equity above the centerline with maximum drawdown worse than the adverse band means the return was purchased with risk the simulation says was not required, and the playbook's instruction is explicit: do not celebrate blindly — audit sizing, overrides, gate compliance, and open exposure. The month is not a failure and the account is not necessarily damaged. What has happened is that the account stepped outside the population it is being measured against, which makes the flattering placement a statement about a different system. The equity number is real. Its provenance is the problem.

Room three

Weak equity, normal drawdown — normal underperformance, and no catch-up.

Below median but above the lower quartile with drawdown behaving is the most common uncomfortable month, and the prescribed response is the hardest: no panic, review expectancy, fees and branch mix, and specifically avoid forced catch-up. The room is uncomfortable precisely because everything is working — the system is producing an ordinary below-centre outcome that a quarter of governed futures produce at any moment, and there is nothing to fix. This is where the strong rule earns its keep, because the instinct to compensate is strongest exactly where compensation has the least justification and the gate ladder has issued no instruction at all.

Room four

Weak equity, adverse drawdown — weak and risky, and the answer is never more risk.

Both placements adverse is the diagnostic room, and the standard's strong rule governs it without qualification: when live performance is below benchmark and drawdown is worse than benchmark, the answer is almost never more risk. It is diagnosis, containment, and a model-to-live mismatch review. Reduce aggression if the gate, throttle, and structural diagnostics support caution — and note the direction of that clause, which permits reduction below what the gate requires but never expansion above it. Where equity sits below the fifth percentile, the room's response escalates further to a formal audit of journal accuracy, branch labels, risk deployment, fees, adherence, and the benchmark's own assumptions.

• The pairing is done before interpretation, not after a conclusion needs supporting.
• Both placements come from their own distributions — equity against the equity table, drawdown against the drawdown table.
• No room authorises deployment beyond the gate cap. The quadrant informs; the ladder governs.

The key idea

One number ranks months; two numbers classify them.

Ranked by equity alone, the four months order themselves in a way that puts the riskiest one first and the cleanest one third. Paired, they separate into four situations with four prescribed responses, only one of which is continue as you were. That reordering is the entire value of the second step, and it costs one additional lookup against a table the workbook already produced. The protocol asks for it every month, in every mood, because the month where it feels least necessary is reliably the month it changes the answer.